One incentive problem is that there are very few IPv6-only services, and, as a matter of policy and practice, people providing services over IPv6 normally don't want to make them IPv6-only. Consequently the people who want IPv6 are all kind of networking geeks; the transition mechanisms like 6over4 and 6to4 and all the rest have really reduced some of the urgency that people might feel!
I mean, it's now easier and less painful, but there isn't any big reason for people to worry about being "left behind"! That may be one reason they don't feel themselves in a hurry.
A world-wide multiprotocol network at the network layer is something that hasn't existed for some years; it's an interesting thing but also a sort of painful thing. In the best case, if your efforts are successful, that is the best we can look forward to for a while!
GJF, I don't think Richard Stallman would appreciate being called a "well known open source evangelist".
In other news
It's funny with what equanimity the reporter covered the story -- it seems she was probably given an assignment to track down all the candidates on the ballot for the office, interview them, and give them all equal space to tell their respective stories in their own words. Accordingly, Daniel gets just as much respect (and space) as Aroner, but there's no news analysis discussing the difference between them.
It seems that Linus Torvalds has written a book, an autobiography, called Just For Fun, forthcoming from HarperCollins.
It also seems that Wells Fargo Bank statements are bearing a notice that
EFFECTIVE 12/01/00, THE FOLLOWING HAS BEEN ADDED TO THE DEPOSIT AGREEMENT: YOUR WELLS FARGO ATM/ATM & CHECK CARD OR INSTANT CASH/CASH & CHECK CARD MUST NOT BE USED FOR ANY UNLAWFUL PURPOSE (FOR EXAMPLE, FUNDING ANY ACCOUNT THAT IS SET UP TO FACILITATE ON-LINE GAMBLING). YOU AGREE YOU WILL NOT USE YOUR CARD OR ACCOUNT FOR ANY TRANSACTION THAT IS ILLEGAL UNDER APPLICABLE LAW.
I wonder why Wells Fargo did that, but it doesn't seem that kelly is around much to comment. It reminds me of some license issues that have come up and that I've been involved in. Of course, bank deposit policies are not exactly open source licenses.
I went to the BayFF meeting at the Moscone center. The meeting was a little scattered. There were three panelists (Scientology critic, reporter, law professor) talking about free speech and intellectual property on-line. I think I might have spotted the Church of Scientology representative. That issue is a bit less good-natured than "Spot the Fed" at DefCon.
Why would you want such a useless thing?
Well: task-no-cleartext-login-daemons, task-no-x11, task-no-gnome, task-no-kde, task-no-lzw, task-no-mp3, (once upon a time) task-no-unlicensed-rsa, task-no-strcpy (or task-no-non-bounds-checking-string-functions; this is problematic because there are safe ways to use these functions and maybe some very important software uses them safely), task-no-games (Don says that's useful for U.S. government procurements). In principle you could have task-no-nonfree (or task-no-semifree and task-no-proprietary), but that's already handled in a different way by Debian (by only shipping free software).
The cool thing is that you can upgrade antipackages to new versions, just like any other packages. It's just that their conflicts will be updated; nothing else will change.
This is a very convenient way to set a policy that a particular category of software should not be installed somewhere. And that policy won't be accidentally overridden by a dependency; instead, the system administrator will be specifically warned about the conflict, and will have to make a choice.
I also have a task-no-outstanding-security-flaws idea, as a way to remove automatically anything with a reported security flaw until its maintainer can fix it (and release a new version, which task-no-outstanding-security-flaws no longer conflicts with). I'm not sure this is an ideal solution, but it's an interesting application of package management.